This was arguably the most momentous week for artificial intelligence since the launch of ChatGPT last year.

OpenAI unveiled the latest version of the technology that underpins its viral ChatGPT chatbot. Elon Musk announced a sarcastic AI ChatGPT rival called Grok coming to his platform, X, formerly known as Twitter. There was a Senate subcommittee hearing about AI regulation in the healthcare space, a targeted attack against OpenAI and the first wearable AI device that intends to one day replace the smartphone.

“These significant announcements are indicative of the speed the AI market is moving at,” said Reece Hayden, an analyst at ABI Research.

Hayden noted the week’s development exemplified what’s unfolding in the industry. AI community continues to balance the risk of unintended consequences by moving too fast while moving as quickly as possible to remain competitive and innovative.

“Overall, it [was] a huge week,” Hayden added.

Here’s a closer look at everything you need to know about the week in AI:

OpenAI hosted its first developer conference about a year after the launch of ChatGPT, which helped renew an arms race among tech companies to develop and deploy similar AI tools in their products.

The company unveiled a series of AI tool updates, including the ability for developers to create custom versions of ChatGPT called GPTs. Similar to plugins, GPTs can connect to databases, be used in emails or facilitate e-commerce orders. CEO Sam Altman, in a few minutes, demonstrated how easy it is for anyone to create a GPT without any prior coding experience.

The company is also rolling out a GPT Store, starting later this month, to allow GPTs to become searchable. Similar to other app stores, they’ll be listed on a leaderboard, and the company will highlight useful tools across categories such as productivity, education and “just for fun.”

Hayden said the announcements, as well as the Apple-like Keynote structure and focus on developers, signaled their intention to “solve their commercial strategic challenges,” which include high costs and limited revenue sources, by building a strong developer ecosystem.

Altman also showed off GPT-4 Turbo, the latest version of the technology that powers ChatGPT. He said it now can support input that’s equal to about 300 pages of a standard book, about 16 times longer than the previous iteration.

Altman shared more about the platform’s growth, too: About 2 million developers now use the platform and about 90% of Fortune 500 companies are using the tools internally. It currently has 100 million active users.

Humane, a startup founded by former Apple employees, introduced its first AI wearable device called the Ai Pin, a small blinking gadget that attaches to clothing. It intends to eventually replace smartphones by projecting information onto a user’s hands, offering the ability to answer phone calls and perform various tasks without holding a smartphone.

The company said it also features a handful of AI-powered tools, including the ability to search, send messages and manage email clutter. The Ai pin, which runs on a Snapdragon processor and boasts a Qualcomm AI engine, is packed with depth and motion sensors, features an ultra-wide camera and a laser ink display.

Arun Chandrasekaran, an analyst at Gartner, said the unveiling marked an “important step in futuristic hardware design for generative and an exploration of a potentially new way of human and machine interaction.”

Still, it’s unclear what adoption could look like. Although the company promises an all-day battery, Hayden said the biggest challenge for on-device AI deployment is battery life. “Given that this device is so small and claims to offer multiple different behaviors and use cases and can sense data to inform the model, it will be interesting to see if the battery life is suitable.”

Other concerns exist, too: “Always-on AI that collects and processes data will really take a leap of faith from society which at present most are unlikely to take,” Hayden said.

The Ai Pin, which starts at $699, will go on sale in the US on Thursday, November 16.

The X app, formerly known as Twitter is seen while a user scrolls through his feed in this photo illustration taken on 10 November, 2023 in Warsaw, Poland. (Photo by Jaap Arriens/NurPhoto via AP)

xAI said in a blog post it took inspiration from The Hitchhiker’s Guide to the Galaxy, a comedy sci-fi novel by British author Douglas Adams. “Grok is designed to answer questions with a bit of wit and has a rebellious streak, so please don’t use it if you hate humor!” xAI said in the post.

Elon Musk’s AI startup xAI unveiled a chatbot called Grok for some users of X, which he suggested has a sarcastic sense of humor similar to his own. Musk, who has owned X — formerly known as Twitter — for a year, said Grok is trained by having “real-time access” to information from the platform.

Musk said Grok is still in the early stages of testing but will soon come to X’s Premium+ service in the US, which includes features such as the blue checkmark for $16 a month.

Musk is a co-founder of OpenAI, but stepped down as chairman five years ago partly over disagreements about the company’s direction.

Two days after the developer conference, OpenAI experienced large-scale outages to its services, which it later attributed to potential targeted attacks on its servers. The company wrote on its website Wednesday evening that it is “dealing with periodic outages due to an abnormal traffic pattern reflective of a DDoS attack.”

A DDoS attack, or distributed denial of service, typically refers to an attacker that floods an internet server to disrupt normal traffic.

Users on Wednesday were unable to access all of OpenAI’s tools and services and received a message that the platform was at capacity.

The company told CNN no user information was compromised.

Major tech companies continue to double down on AI, too. Reuters reported Amazon is investing millions of dollars on training an AI codenamed “Olympus,” which is expected to have double the amount of “parameters,” or building blocks that make AI smarter than OpenAI’s GPT-4 model

In addition, YouTube is testing AI tools that answers questions about content, makes recommendations and summarizes topics in the comment sections of videos.

Chandrasekaran said that while not all companies will create huge AI models, many will continue to build smaller and specific ones to improve their products, automate tasks and gain a competitive edge.

– CNN’s Anna Coobin contributed to this report



Read the full article here

Share.
Exit mobile version